iPTF16geu: Lensed Type Ia Supernova

• iPTF16geu is a multiply imaged, strongly lensed Type Ia supernova at z=0.409, displaying four images in an Einstein-cross configuration.
• The system exhibits high magnification (up to ~68) and extremely short time delays (<1 day), challenging conventional cosmographic methods.
• Advanced observations with adaptive optics and HST have enabled precise separation of the supernova, host, and lens light, refining lens models and microlensing diagnostics.

Searching arXiv for recent and foundational papers on iPTF16geu. iPTF16geu is a multiply imaged, strongly gravitationally lensed Type Ia supernova at redshift $z=0.409$, lensed by a foreground galaxy at $z=0.2163$ into four images arranged in an Einstein-cross configuration around the deflector. Discovered in near real time by the intermediate Palomar Transient Factory and rapidly followed with adaptive optics and the Hubble Space Telescope (HST), it became the first resolved strongly lensed Type Ia supernova and a benchmark system for combining standardizable-candle photometry, time delays, lens-galaxy dynamics, and microlensing diagnostics on physical scales of $\sim 1$ kiloparsec (Goobar et al., 2016, Lee, 2017).

1. Discovery and basic configuration

iPTF16geu was discovered by Ariel Goobar and collaborators in the intermediate Palomar Transient Factory survey as an unusually bright Type Ia supernova for its redshift. Spectroscopic classification established it as a normal SN Ia, and the transient appeared $4.3$ magnitudes brighter than expected for a standard SN Ia at $z=0.409$, implying a magnification of $\sim 52$ in flux. High-spatial-resolution observations with Keck adaptive optics, VLT/NACO, and HST resolved four distinct images at radii $0.2679$–$0.3133$ arcsec from the lens center, with approximately $90^\circ$ azimuthal separations (Goobar et al., 2016).

The resolved geometry established iPTF16geu as a quad lens with close alignment between the line of sight to the supernova and the lens galaxy. The event was also unusual because all key lensed images could be observed in the fading phase. That combination made the system immediately relevant to two distinct programs: direct cosmography with a lensed standard candle, and empirical tests of small-scale structure in lens galaxies through image-by-image magnification anomalies (Lee, 2017).

2. Refined photometric measurements and empirical observables

Subsequent HST and Keck adaptive-optics reference imaging obtained after the supernova faded enabled a direct separation of supernova, host, and lens light. On that basis, the total lensing magnification was measured as

$\mu = 67.8^{+2.6}_{-2.9},$

with extinction corrections applied for both host and lens galaxy dust. The corresponding extinction-corrected image magnifications were reported as $z=0.2163$0, $z=0.2163$1, $z=0.2163$2, and $z=0.2163$3 mag. In the same analysis, the host reddening was inferred as $z=0.2163$4 mag, while the lens reddening varied strongly by image, with $z=0.2163$5 values of $z=0.2163$6, $z=0.2163$7, $z=0.2163$8, and $z=0.2163$9 mag for images 1–4. The inferred total magnification was found to be insensitive to assumptions about the dust properties in the host and lens galaxy (Dhawan et al., 2019).

The same photometric study measured very short delays for the three fainter images relative to the brightest one: $\sim 1$0 These values are consistent with the near-simultaneity already anticipated from the lens symmetry. A separate spectroscopic analysis of the two brightest images obtained from resolved HST slitless spectra yielded $\sim 1$1 days, consistent with the photometric estimate and with a negligible delay for most practical purposes (Johansson et al., 2020).

3. Time-delay cosmography and the standard-candle advantage

The principal cosmological interest of iPTF16geu lay in the prospect of measuring the Hubble constant through lensed-supernova time delays. For a multiply imaged transient, the delay between two images may be written as

$\sim 1$2

where the time-delay distance $\sim 1$3 is proportional to $\sim 1$4. In a Type Ia supernova, the intrinsic luminosity is standardizable, so the absolute magnification can also be estimated: $\sim 1$5 That extra observable is the key distinction from quasar lenses, because it can be used to break the mass-sheet degeneracy that otherwise limits $\sim 1$6 inference from time delays alone (Lee, 2017).

For iPTF16geu, however, the same compact and symmetric lens configuration that produced a spectacular Einstein cross also imposed severe practical limitations. Lens models constrained by HST imaging predicted maximum relative delays of less than one day, with image B arriving first and the other three images following within $\sim 1$7–$\sim 1$8 days; the discovery paper had already quoted all delays as less than $\sim 1$9 hours. Such short delays require very high cadence and very stable treatment of microlensing and dust to become cosmographically decisive (More et al., 2016, Goobar et al., 2016).

Microlensing studies sharpened that limitation. Simulations showed that lensed SN Ia color curves are nearly achromatic during the first $4.3$0 rest-frame weeks after explosion, making early multiband follow-up especially valuable; in that regime, microlensing-induced time-delay errors can be strongly reduced. For iPTF16geu specifically, because most HST follow-up occurred in the chromatic phase and the intrinsic delays were only $4.3$1–$4.3$2 days, the forecast microlensing-induced uncertainty was $4.3$3, dropping to $4.3$4 only in the hypothetical case of earlier achromatic-phase color-curve coverage (Goldstein et al., 2017). Later population studies therefore treated iPTF16geu as a proof of principle for lensed-SN cosmography, but not as a favorable cosmographic system in itself; its short delays make it unsuitable for precision $4.3$5 work compared with the longer-delay systems expected in deeper surveys (Murieta et al., 2023).

4. Flux-ratio anomalies, microlensing, and lens-galaxy structure

The central modeling problem posed by iPTF16geu is that smooth macrolens models reproduce the image positions far more easily than the image fluxes. The original lens models predicted broadly similar magnifications for the four images, yet the observed flux ratios differed strongly; in the discovery analysis, image 4 was predicted to be the brightest but was observed to be the faintest, and the relative magnifications were taken as evidence for sub-structures in the lensing galaxy (Goobar et al., 2016). A later lens-modeling study formalized the same conclusion: a smooth lens density fails to explain the iPTF16geu fluxes regardless of the slope, and the total probability for a smooth halo model combined with star microlensing to explain the observed image fluxes is maximized at $4.3$6 for $4.3$7, with a corresponding lower bound $4.3$8 at $4.3$9 confidence level (Mörtsell et al., 2019).

Interpretation of the flux anomalies has remained non-unique. One line of analysis argued that the discrepancies were too large to be due to microlensing alone: Monte Carlo experiments gave probabilities as low as $z=0.409$0 for a “typical SN Ia” case and $z=0.409$1 for an “atypical SN Ia” case, rising only to $z=0.409$2 for certain macro-model variations, while microlensing broadened the inferred intrinsic source magnitude to a full width half maximum of $z=0.409$3 magnitudes (Yahalomi et al., 2017). A related study concluded that image A is nearly $z=0.409$4 magnitudes brighter than macromodel expectations and that iPTF16geu is not a “standardisable” configuration in the sense of having microlensing scatter below $z=0.409$5 mag; by that criterion, symmetric small-Einstein-radius systems such as iPTF16geu are disfavored for precision standard-candle work (Foxley-Marrable et al., 2018).

A second line of work emphasized that the anomaly depends sensitively on the adopted macro model. Two-component baryonic mass models were introduced to remove the need for the shallow projected density slope $z=0.409$6, the $z=0.409$7 kpc offset between mass and light centroids, and the $z=0.409$8 mass-light misalignment that arose in earlier single-component fits. In these models, the total mass distribution remains single-centered but becomes lopsided and non-self-similar with radius, reducing the required microlensing to $z=0.409$9–$\sim 52$0 magnitudes for three of the images while reproducing the host-galaxy ring morphology more naturally (Williams et al., 2020).

A third line of work combined image positions, magnifications, and the lack of strong fluctuations in the four light curves. In a baryon-plus-dark-matter macromodel with stellar microlenses drawn from the inferred baryonic component, larger stellar surface mass densities were preferred after marginalization over macro and micro parameters, and the authors concluded that microlensing from the baryonic component suffices to explain the observed flux ratios and light curves (Diego et al., 2021). More recently, joint inference from microlensing magnification maps and the standardizable-candle constraint yielded a lens mass slope $\sim 52$1 for iPTF16geu and no evidence for a population of dark compact objects, with a $\sim 52$2 upper limit $\sim 52$3 for compact-object masses $\sim 52$4 (Arendse et al., 2 Jan 2025). A point-source-only forward-modeling pipeline based on GIGA-Lens later found $\sim 52$5, fully consistent with published mass-parameter inferences despite not using the extended host-galaxy images (Baltasar et al., 26 Jan 2026).

Taken together, these analyses make one point unambiguous: the image fluxes of iPTF16geu cannot be treated as a straightforward product of a smooth elliptical macromodel. What remains debated is the exact partition of responsibility among stellar microlensing, lens-galaxy baryonic complexity, and other substructure.

5. Spectroscopy, supernova classification, and explosion-physics constraints

Spectroscopically, iPTF16geu rapidly became more than a lensing case study. Ground-based spectroscopy, amplified by a factor of $\sim 52$6 from lensing magnification, enabled unusually high signal-to-noise observations of a high-redshift SN Ia, and HST slitless spectroscopy resolved spectra of individual lensed supernova images for the first time. The spectroscopic campaign measured host and lens redshifts of $\sim 52$7 and $\sim 52$8, and refined the lens-galaxy line-of-sight velocity dispersion to $\sim 52$9, a quantity of direct relevance to lens-mass modeling (Johansson et al., 2020).

The supernova itself was consistently found to be spectroscopically normal. Pseudo-equivalent widths of Ca II H&K, Mg II, Fe II, Si II $0.2679$0, and the Ca II IR feature followed the evolution seen in local comparison samples, and the Si II $0.2679$1 expansion velocity at $0.2679$2 was reported as $0.2679$3 with a velocity gradient $0.2679$4, placing the event in the high-velocity but not high-velocity-gradient subclass under the Folatelli/Branch–Wang criteria. The same study also reported strong Na I D absorption in the host, with total equivalent width $0.2679$5 Å decreasing to $0.2679$6 Å over 9 days, implying $0.2679$7–$0.2679$8 mag from Na I D despite much lower extinction from light-curve fits, $0.2679$9–$0.3133$0. That combination placed iPTF16geu among SNe Ia with anomalously large Na I D absorption (Johansson et al., 2020).

Later comparisons with other lensed SNe Ia extended the spectroscopic significance of the event. Using the standard Doppler estimator

$0.3133$1

the Si II $0.3133$2 Å and Ca II NIR triplet velocities of iPTF16geu were found to be compatible with the low-$0.3133$3 SN Ia distribution, showing no sign of evolution with redshift and supporting the broader claim that SN Ia spectroscopic properties remain stable across cosmic time at least out to the regime probed by lensed events (Dhawan et al., 2024). A subsequent comparison to radiative-transfer explosion models identified the delayed-detonation models DDC6 and PDDEL1 as the closest overall matches to the light curves and many absorption features, while also finding that iPTF16geu is systematically brighter than all models in the observer-frame $0.3133$4 band by $0.3133$5 mag and exhibits a rest-frame NUV flux excess, with $0.3133$6, above all tested models. No combination of magnification $0.3133$7 and reddening $0.3133$8 reconciled all observed properties, especially in the UV (Murieta et al., 1 Oct 2025).

6. Survey selection, population context, and continuing legacy

iPTF16geu is now understood as an archetype of what shallow time-domain surveys preferentially discover. In pre-LSST surveys, the effective strategy is magnification selection: transients are first noticed because they are too bright for their host-galaxy redshift, and only later confirmed as lenses by high-resolution imaging. Monte Carlo calculations showed that iPTF16geu, with $0.3133$9 and $90^\circ$0, lies within the $90^\circ$1 contours of predicted redshift–magnification space for iPTF-class surveys, so its discovery does not require abandoning standard expectations for shallow-survey selection (Wojtak et al., 2019).

More detailed population studies sharpened that interpretation. Simulations that include microlensing and realistic survey depths showed that shallow surveys such as iPTF and ZTF are biased toward highly magnified quad systems with short time delays and small angular separations; iPTF16geu and SN Zwicky are therefore consistent with the discovery space of shallow surveys, but not representative of the longer-delay, larger-separation population expected from deeper surveys. In one such forecast, LSST was predicted to find $90^\circ$2 lensed supernova systems in ten years, with $90^\circ$3 suited for cosmography and enabling a $90^\circ$4 precision $90^\circ$5 measurement with LSST glSNe (Murieta et al., 2023).

Realistic ZTF simulations also clarified why events of the iPTF16geu class remain rare in practice. When the Bright Transient Survey magnitude cut $90^\circ$6 mag is imposed, the expected number of identifiable lensed SNe Ia drops from $90^\circ$7 per year to $90^\circ$8 per year, strongly biasing the sample toward extreme magnification. Under that selection, highly magnified systems such as iPTF16geu and SN Zwicky are not anomalous outliers but the ones most likely to be recognized (Carracedo et al., 2024).

The system has also entered spectroscopic lens-selection work. The DESI Single Fiber Lens Search independently recovered the host galaxy of iPTF16geu from a background [O II] doublet in the foreground luminous red galaxy spectrum, demonstrating that compact galaxy-scale lenses with Einstein radii of order $90^\circ$9 can be identified spectroscopically even without the transient. That result is significant because monitoring such spectroscopically selected lens catalogues creates a direct path to future multiply lensed transients, time-delay cosmography, and flux-ratio measurements of substructure (Karp et al., 3 Dec 2025).

In that sense, iPTF16geu occupies two distinct historical positions. It is both the first resolved strongly lensed Type Ia supernova and a limiting case that exposed, with unusual clarity, the practical interaction between standard-candle magnification, sub-day delays, dust, stellar microlensing, and lens-galaxy complexity. Its enduring importance lies less in the precision $\mu = 67.8^{+2.6}_{-2.9},$0 measurement that was initially envisioned than in having defined the observational and modeling problems that now shape the study of lensed supernovae.